Optimization of Session Topic Backtracking Method Based on Interactive Text

Conversation topic tracing refers to the use of conversation topics to trace back the subjects and related information associated with the topic. This article first focuses on the need for topic backtracking and constructs a conversation topic backtracking model. Secondly, optimize the feature selection algorithm TFIDF and propose an improved algorithm LRIDF. Thirdly, applying the above research results, propose an optimized interactive text conversation topic backtracking method. Empirical analysis shows that this method can effectively improve the performance of conversation topic backtracking.